www.ijcrsee.com
823
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
Review Article
Received: August 14, 2025.
Revised: November 20, 2025.
Accepted: December 01, 2025.
UDC:
323.23:004.8
316.77:004.8
10.23947/2334-8496-2025-13-3-823-835
© 2025 by the authors. This article is an open access article distributed under the terms and conditions of the
Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
*
Corresponding author:
radoslav@diplomacy.bg.ac.rs
Abstract: Since artificial intelligence (AI) has been integrated into our digital communication landscape, there have
been major changes in how political campaigns are strategically designed and how public opinion is influenced. With the help of
machine learning (ML), generative models (GM) and natural language processing (NLP), AI tools have introduced new opportuni-
ties for political engagement. Today, thanks to AI-driven data analytics, we can micro-target voters based on their psychographic
profiles and adapt political messages with incredible precision. On the other hand, generative AI technologies are increasingly
used to spread false information or to imitate political endorsements, which has a great impact on public opinion. The dissemina-
tion of such content can greatly reinforce ideological prejudices and contribute to social divisions. This paper draws on recent
empirical research and case studies to illustrate how AI-generated disinformation campaigns can affect electoral processes
and undermine trust in democratic institutions. Various examples, such as the use of bots to control social media to deepfake
content impersonating political figures, show that ethical, technological and legal safeguards are urgently needed. Furthermore,
this paper supports an approach to AI governance that strikes a balance between promoting innovation and reducing harm. This
implies the development of tools for AI detection, transparency measures and cooperation between sectors in order to promote
responsibility and integrity of information. Greater digital literacy among citizens and proactive policy responses will be necessary
in the near future to ensure the resilience of democratic systems due to the increasingly rapid development of AI technology.
Keywords: Artificial Intelligence (AI), Political Communication, Generative AI, Deepfakes, Disinformation, Large
Language Models (LLMs).
Radoslav Baltezarević
1*
, Vladimir Lović
2
, Ivana Baltezarević
3
1
Institute of International Politics and Economics, Belgrade, Serbia, e-mail:
radoslav@diplomacy.bg.ac.rs
2
International center for economics and public policy - ICEPP, Belgrade, Serbia, e-mail:
vladalovic@yahoo.com
3
Institute for Political Studies, Belgrade, Serbia, e-mail:
ivana.baltezarevic@ips.ac.rs
Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication
Introduction
According to Fetzer and Fetzer (1990), artificial intelligence (AI) is the ability of computers to imitate
human intelligence and to think and learn in the same ways as human beings. The impact of AI, due to
its rapid development, is becoming greater in numerous fields, especially in politics. The rise of digital
technology has brought a host of highly complex challenges to the sharing, reliability and consumption of
political information. This shift prompts a careful examination of the role of AI in political communication and
highlights the need for strategies that promote greater transparency and, in our digital age, foster informed
public discussions (Bareis and Katzenbach, 2022).
The acceptance of cultural values and public opinion are greatly influenced by the media, many of
which are deceptively promoted as independent ideas (Baltezarević et al., 2014). However, the rise of new
technologies is increasingly suppressing traditional cultural values and narratives, often replacing them
with algorithmically generated content (Baltezarević et al., 2019). AI has revolutionized the way we analyze
extensive data sets, including demographic, behavioral, and psychological data derived from individuals’
online behaviors, as well as the growing issue of deepfakes (Chester and Montgomery, 2017). These tools
are often used to create persuasive political messages (Carr, 2011).The emergence of “deepfakes,” that is,
realistic fake videos produced through AI-based facial manipulation, has raised social alarms due to their
www.ijcrsee.com
824
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
ability to imitate a person’s actions or speech with minimal signs of manipulation (Chawla, 2019).
Alexander Nix (the former CEO of Cambridge Analytica) mentioned that by understanding the psy-
chological characteristics of target audiences, political communications can be adjusted to connect with
specific traits, be it emotional, rational, or fear-driven (Mermoud, 2017). Large language models (LLMs)
are typically fine-tuned to boost user engagement by spotting patterns that create emotionally resonant
and engaging content (Ethayarajh et al., 2024). A focus on engagement can often amplify controversial
narratives, as people tend to connect with content that is consistent with their beliefs (Brady et al., 2017).
More than half of Americans voiced serious concerns about the spread of AI-generated political
propaganda, according to a 2023 survey, while 23% were only moderately concerned (Statista Research
Department, 2024). These worries are not unfounded, as there have already been several reports of AI
being abused in political communication. For instance, the founder of the investigative website Bellingcat,
Eliot Higgins, created fake photos of Donald Trump’s arrest using the Midjourney AI platform, and the
German political party AfD used AI-generated graphics to inflame anti-refugee sentiment (Matheis, 2023).
According to a recent survey, social media is one of the main sources of news for 56% of respond-
ents, but almost 70% of them believe that it is also the largest source of false information, raising serious
questions about its reliability (Weforum, 2024). This dynamic, with the advent of generative artificial intel-
ligence (GenAI), only adds to the complexity. GenAI systems can both create and identify fake content
(Loth et al., 2024). For example, OpenAI’s GPT models are able to produce content that closely resembles
the style, tone, and structure of reliable news sources, making it difficult to distinguish between authentic
journalism and made-up stories (Brown et al., 2020).
To tackle this challenge, technology and human oversight must collaborate. Large language models
(LLMs) can indeed spread false information, but they also have the potential to help us fight against it (as-
suming we implement human intervention effectively). It’s vital for users, content moderators, and news
organizations to take proactive measures to curb the dissemination of misleading information. Reporting
questionable content can enhance AI detection systems and help quickly identify harmful or deceptive
information (Virginia Tech, 2024).
On a broader level, AI has significant implications for democratic processes. The ability of AI to
create misleading information, influence voter behavior, and potentially undermine election integrity is a
major concern (Coeckelbergh, 2022). Yu, 2024 in their study highlights the necessity of finding a balance
between robust regulatory measures, technological advancements, and ethical oversight to mitigate these
risks. With the potential to distort political realities and threaten democratic institutions, deepfakes and AI-
generated misinformation are among the most pressing threats to the integrity of elections.
Ultimately, it’s essential to reflect on the ethical implications and the potential of AI technologies as
they advance. Since AI is still in the early days of development, continuous research, development, and
critical evaluation are key to solving current issues and shaping the future of this technology in a way that
bolsters democratic processes instead of undermining them (Baltezarević and Baltezarević, 2024).
In order to explore this dynamic more thoroughly, this paper is guided by the following research questions:
1. In what way does artificial intelligence (AI) influence and change strategies in the field of political
communication in modern democratic societies?
2. How does AI-powered voter profiling influence the shaping of public opinion?
3. What ethical and legal issues emerge from the political use of deepfakes and AI-generated misinformation?
4. How do AI algorithms lead to political polarization on social media by creating filter bubbles and echo
chambers?
5. What legal and technological measures can be implemented to minimize the dangers of AI misuse
while still upholding free speech and democratic values?
The Role of AI in Political Communication, Campaign Strategy, and Public Opinion Manipulation
AI is defined as any computer or algorithm that can observe its surroundings, learn from them,
and make smart decisions based on that knowledge. This definition is quite expansive and includes a
variety of technologies, but machine learning (ML) techniques are currently some of the most popular
approaches (
Samoili et al., 2020).
AI is dramatically changing the way politicians connect with the public and shape opinions. Its abil-
www.ijcrsee.com
825
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
ity to analyze complex data, process information, and adjust communication strategies based on user
preferences makes it a game-changer in modern politics (Crawford, 2021). One of the key applications of
AI in this field is data-driven voter profiling, where potential voters are linked to personality types through
AI-powered analytics. When we classify personality types using psychographic and demographic fac-
tors, we can make informed predictions about how people will respond to different stimuli. This skill gives
campaigns a crucial advantage, helping them to sway swing voters and inspire their target audience to
participate in elections (Wakefield, 2019).
To make sense of the huge amounts of data flowing from social media and online platforms, politi-
cians are increasingly turning to AI-based analytics to get a clearer picture of what voters really think and
feel (Battista and Uva, 2023). This growing reliance on AI tools is enhancing election campaigns and com-
munication strategies, making them more targeted and effective (Alvarez et al., 2023). AI also allows for
real-time monitoring of public sentiment on political issues, debates, and candidates. Campaigns can lev-
erage this data to identify potential voters and tailor personalized ads that either encourage or discourage
them from voting. Consequently, political leaders can adjust their positions and communication strategies
to keep up with the constantly shifting landscape of online public opinion (Lutkevich and Hildreth, 2022).
AI is revolutionizing campaign management by not only boosting financial contributions but also
enhancing how messages are crafted. A striking example of this is the “Vote Leave” campaign during the
Brexit referendum, which cleverly used A/B testing to tailor their messages for specific demographic seg-
ments. Campaign director Dominic Cummings explained that the catchphrase “Let’s take back control”
was deeply rooted in a careful analysis of public opinion on the European Union (EU). Through their
iterative testing, they found that the word “back” in the slogan sparked anger by triggering loss aversion,
the psychological tendency to prefer avoiding losses over acquiring equivalent gains, especially when it
comes to control (Schneider, 2017).
To more effectively reach and engage their target voters, candidates and their campaign staff
can harness the predictive power of AI by combining data from a variety of sources (Crilley, 2018). This
capability allows for the development of highly customized political messages that resonate with voters’
unique needs and preferences, enhancing the campaign’s overall impact (Nunziata, 2021). The use of AI
is growing rapidly, especially for real-time fact-checking and analyzing public reactions to political events.
For example, Microsoft’s AffectiveSpotlight employs head movements and facial expressions to assess
viewers’ emotional responses, allowing presenters to tweak their delivery style (Murali et al., 2021).
Political debates can benefit greatly from AudienceView, which uses large language models (LLMs)
to classify audience feedback, thus directly helping journalists better understand public opinion (Brannon
et al., 2024). In addition to this system, Factiverse and Full Fact are real-time fact-checking websites that
have introduced AI systems capable of capturing, evaluating, and analyzing political commentary in real
time (Corney et al., 2024). A billion web pages may seem impossible to scan. However, Factiverse is mak-
ing it feasible for both companies and individuals to extract the information they require into essential, easily
assimilated insights using intelligent prioritization and robust crawling tools (Factiverse, 2025). Full Fact is
a British nonprofit that verifies and corrects news reports and statements that make the rounds on social
media (Dudfield, 2025). Full Fact AI uses cutting-edge artificial intelligence to enable fact checkers, journal-
ists, researchers, and communicators to recognise, validate, and refute misleading content (Fullfact, 2025).
AI’s impact stretches far beyond just campaigning; it’s also leaving its mark on governance. One of
the standout advantages is its knack for sifting through massive amounts of data to predict the outcomes
of political decisions. AI has also changed how people communicate with their representatives, introduc-
ing fascinating new features like chatbots and virtual assistants on company websites and AI-powered
social media interactions. This has opened up more direct communication channels between elected of-
ficials and the people they represent (
Viudes, 2023). Still, there are some risks to consider. For instance,
during the 2025 protests in Los Angeles, AI chatbots such as Grok and ChatGPT were found to spread
false information because they misinterpreted images and lacked the necessary context (Gilbert, 2025).
AI chatbots can provide voters with tailored insights about candidates and policies, which can boost
participation and support informed decision-making (
Political Communication, 2023). However, the rise
of AI-generated content, including deepfakes and synthetic media, poses a serious risk to the integrity of
political communication (
Thornhill, 2024). These tools can be misused to spread false information, influ-
ence public opinion, and erode trust in democratic institutions, particularly in nations with lower media
literacy levels (Funk et al., 2023). To truly harness the benefits of AI, it’s essential for governments, civil
www.ijcrsee.com
826
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
society, and tech companies to collaborate and create strategies that are specifically designed for differ-
ent contexts (Hagerty and Rubinov, 2019).
We tend to underestimate the importance of improving our digital literacy, especially given the threats
posed by deepfakes and AI-generated disinformation. The speed at which synthetic media is evolving is
outpacing our technology and legal systems. Encouraging media literacy and critical thinking skills has
become a more effective and scalable way to address these problems. Recent research shows that digital
literacy programs can successfully reduce people’s vulnerability to false information by improving their
ability to spot content that is manipulative and better understand how certain narratives are promoted by
algorithmic systems (Guess et al., 2020). It is also crucial that both governments and digital platforms work
together to promote public education through accessible initiatives, such as media warnings, verified fact-
checking labels, and very clear explanations of algorithms (Roozenbeek and van der Linden, 2020). Even
the most sophisticated detection technologies can become useless if there is no widespread digital literacy,
as end users remain vulnerable to emotionally charged synthetic content (Sustainability Directory, 2025).
AI-driven bots have become quite common on social media platforms. Typically powered by ML
algorithms, these bots can imitate human behavior at an impressive speed while still appearing genuine.
They use natural pauses and responses to create the illusion that a human is behind the messages (Bessi
and Ferrara, 2016). Any government or political party with enough financial backing can deploy a whole
army of social bots to influence public discussions on social media, removing the need for specialized
technical skills and resources (Ferrara et al., 2016).
While these technologies can enhance social interactions and deepen our understanding of one
another, they also come with risks like deceit, manipulation, and the spread of misinformation (Gallo et
al., 2022). The influence of social bots is amplified by how easy they are to create and manage. Social
bots often take on tasks such as researching hashtags and keywords, posting content, responding to user
interactions, following users interested in specific topics, and gathering opinions on online discussions
(Ferrara et al., 2016). The misinformation generated by AI, especially the biased or fake news produced by
advanced text generation systems, could seriously undermine political engagement and pose a threat to
democratic processes (Klinger et al., 2023).
We can theoretically distinguish between “good” and “bad” social bots. Bad (malicious) bots usually
spread harmful links or misleading stories, while good bots share posts that offer valuable information. Bad
bots often create fake accounts that look almost identical to real users, making it tough to differentiate them
from trustworthy content. As noted by BSI (2025), these bots typically engage in coordinated “fake news”
efforts that seek to shape public opinion. Pamment et al. (2018) highlight that this dishonest strategy can
significantly impact political backing and alter our perception of reality in multiple ways.
The growing presence of bots is only widening the gap in society, creating confusion, and shaking
public trust in information. This makes it increasingly tough to tell what’s real and what’s not (Bradshaw and
Howard, 2018). For the first time in over ten years, automated traffic has overtaken human internet usage,
making up almost 51% of all web traffic, as highlighted in the 2025 Imperva Bad Bot Report. The surge is
mainly driven by large language models (LLMs) and the swift rollout of AI technology (Chang, 2025).
AI is revolutionizing fields such as speech and sentiment analysis, as well as content creation and
distribution. To grasp the emotional tone and reactions of the public, political campaigns can use sophisti-
cated algorithms that analyze speeches and communications, helping them detect how audiences are feel-
ing (Khare, 2023). Sentiment analysis is a vital part of natural language processing (NLP) and categorizes
emotional content in text as neutral, negative, or positive (based on subjective data) (Rajashekhargouda,
2022
). This analytical approach improves business intelligence and provides measurable insights that can
improve strategic decisions (Kumar and Garg, 2020). Sentiment analysis models can reveal the polarity of
opinions, the topics being discussed, and the individuals behind them (Obot et al., 2025).
AI is also influencing social media influencer marketing strategies outside of the political campaign
space. Technologies created by a number of advertising agencies assess an influencer’s “brand safety”
and forecast whether or not they will participate in political discourse. A marketing company called Captiv8,
which works with companies like Kraft Heinz, for example, recently introduced an AI-powered tool that
analyses social media users’ online mentions to determine how likely they are to discuss elections or other
politically sensitive topics. In this system, an “A” means you should exercise extreme caution, while a “C”
suggests a profile that’s generally safe. Influencers receive these letter ratings based on their writing, com-
ments, and how they’re covered in the media. These ratings consider delicate matters like hate speech,
www.ijcrsee.com
827
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
violence, sexual content, and divisive social issues (Maheshwari, 2024)
.
Artificial Deception: The Crisis of Democratic Integrity, Deepfakes, and Disinformation
The rise of AI brings with it some crucial ethical dilemmas, particularly around the accessibility and
transparency of information (Nida-Rumelin and Weidenfeld, 2019). With AI technologies impacting every
facet of how we produce and share information, these challenges are becoming more pressing in our rap-
idly changing digital world. One of the most concerning issues we are facing is the rise in deepfakes. This
not only threatens the integrity of our political systems but also puts our society’s trust and stability at risk
(Westerlund, 2019).
Deepfakes are made by combining deep learning techniques with altered content. This technology
creates a deceptive reality that can be difficult to detect with just a glance, allowing a person’s face to
convincingly express emotions or say things they’ve never actually uttered (Korshunov, 2018). Thorough
training, the generator (Generative Adversarial Networks - GANs), enhances its ability to create more re-
alistic fake images, driven by the competitive dynamics of the system (Shen, 2018). This has resulted to
deepfakes becoming increasingly intricate and more difficult to spot.
A diverse array of groups, including governments, political activists, criminals, dubious individuals,
conspiracy theorists, and automated bots, are leveraging deepfakes to achieve their goals. According to
Zannettou (2019), these objectives can differ greatly, ranging from influencing public sentiment and inciting
social chaos to seeking financial gain and demonstrating ideological loyalty. In the political arena, deepfakes
are particularly worrisome. They can produce fake videos of people appearing to say or do things they never
actually did. This can confuse voters and might even change the course of elections. Such manipulation is a
serious threat to democracy, as it prevents individuals from making well-informed choices (Cheguri, 2023).
We must also be aware of the ways these technologies can be misused, such as in financial fraud,
scams, hoaxes, fake news, non-consensual pornography, extortion, harassment, bullying, electoral inter-
ference, and the spread of disinformation. However, although deepfakes are often viewed negatively, they
can actually be quite innovative and beneficial. They have the potential to vividly reenact historical events,
enhancing the authenticity of films or serving as useful educational resources (Cruci, 2024). With the rapid
advancement of deepfake creation methods and the development of AI detection tools to counter these
risks, it’s more important than ever for people to learn how to identify misleading media (Appel and Prietzel,
2022). The very AI technologies that make it possible to produce convincing deepfakes also power the
tools used to identify them, resulting in a dynamic and ever-changing battleground between creators and
defenders. However, this leads to an ongoing technological arms race.
The public’s trust in reliable sources of information and organizations has really taken a hit due to
these changes. Deepfakes, being trickier to spot than traditional fake news, only add to the cybersecurity
challenges that both individuals and organizations face. Plus, since deepfakes blur the lines between real-
ity and fiction, they often have a stronger impact on shaping public opinion (Hu et al., 2022). For instance,
Google Trends data reveals that searches for “free voice cloning software” skyrocketed by an astonishing
120% from July 2023 to 2024 (Cruz, 2024), highlighting how accessible technology has made it for even
amateurs to create altered audio recordings.
Deepfakes are captivating and novel, which helps them spread false information at lightning speed.
Studies indicate that the top 1% of rumors usually don’t get more than 1,000 views on platforms like Twitter
(now X). In fact, even true news stories tend to have a smaller reach, illustrating how misinformation can
travel faster and further than the truth (Vosoughi et al., 2018). To make things even trickier, the technology
behind deepfakes is advancing at a pace that outstrips our current detection methods (Surfshark, 2025).
It’s difficult to overlook the impact that AI and deepfakes are having on elections. A 2024 survey
found that more than 75% of people globally are anxious about AI’s potential impact on future elections.
The United States and Singapore are particularly concerned, with 72% and 83% of respondents, respec-
tively, voicing their worries (Petrosyan, 2025). Furthermore, large language models (LLMs) and AI tools like
Midjourney, Google’s Gemini, and OpenAI’s ChatGPT have made it easier to normalize deepfakes across
multiple platforms (Zandt, 2024).
According to data collected since 2017, 31% of deepfake cases have involved fraud, with particular
focus areas being celebrities, and political figures. The targeting landscape shows that 35% of incidents
were aimed at politicians or celebrities, while a significant 65% impacted the general public. Of these in-
www.ijcrsee.com
828
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
stances, 27% were politically provocative, and 25% contained explicit content. US President Donald Trump
was the most affected, being the target of 18% of the deepfakes involving politicians, which amounts to 25
incidents. Joe Biden faced 20 deepfakes, including voice robocalls that were manipulated and often linked
to election matters. Additionally, well-known politicians like Kamala Harris and Volodymyr Zelenskyy also
experienced multiple attacks (Surfshark, 2025).
Additionally, large companies are becoming more susceptible to crimes made possible by deep-
fakes. For example, using a deepfake impersonation of the CFO of the British engineering company Arup,
fraudsters were able to successfully transfer $25 million to Hong Kong bank accounts in 2024. The scam
was confirmed by the deceitful CFO and team members who participated in video calls (Noto, 2024). An-
other significant case occurred in 2019, when fraudsters used a deepfake voice of the president of a UK
energy company to siphon off €220,000 (Somers, 2020).
A low-tech deepfake video that purported to show American politician Nancy Pelosi intoxicated went
viral in 2019, deceiving many Facebook users with almost 2.5 million views (CBS News, 2019). Globally,
political deepfakes have become more complex. In early 2024, robocalls featuring a deepfake voice of
President Biden were sent to thousands of voters in New Hampshire, discouraging them from voting in the
primary election. The fact that these audio recordings were created in less than 20 minutes for just $1 each
shows how readily and affordably such false information may propagate (Seitz-Wald, 2024). During the
2023 annual news conference with Russian President Vladimir Putin, a student from St. Petersburg caught
everyone’s attention by asking a question. What made this moment particularly striking was the use of
deepfake technology. The voice and appearance of this individual were actually created using a deepfake
AI avatar of Putin himself (NBC News, 2023).
European laws could serve as a useful blueprint for tackling the issue of false information generated
by AI. The EU’s Digital Services Act requires tech platforms to evaluate the risks their products pose to
society, particularly concerning democracy and elections. Moreover, these platforms are required to pro-
vide relevant data to independent experts to help evaluate their impacts (Hetrick, 2024). As policymakers
around the world tackle the rapidly changing landscape of AI technologies, these regulatory frameworks
are working to find a balance between the risks and rewards associated with AI.
On a global scale, we’re starting to see legislative measures take shape in response to these chal-
lenges. To boost transparency, the European Union’s AI Act of 2024 requires that AI-generated content be
clearly labelled. At the same time, several states in the US have enacted laws that make it illegal to create
and distribute harmful deepfakes, particularly those intended to influence elections or spread misinforma-
tion (Kumar, 2025).
Regulating AI in political communication raises a number of legal concerns, particularly with regard
to platform liability, foreign jurisdiction, and the tension between free speech and censorship. The European
Union’s AI Act and Digital Services Act (DSA) aim to regulate AI use and hold online platforms accountable
(European Parliament, 2022). These laws prohibit the use of manipulative AI and also require platforms to
eliminate illegal or harmful content, like hate speech and misinformation, by specific deadlines (European
Commission, 2021). Although these strategies are designed to protect users, critics warn that they might
lead to the over-removal of lawful content, which could violate free speech rights (Lazaro Cabrera, 2024).
With the rise of deepfakes, misinformation, often referred to as “fake news”, has become alarmingly
prevalent on social media in recent years. This trend has had a profound effect on our economy, politics,
democracy, and society as a whole (Burkhardt, 2017). Fake news is often fueled by financial or political
motives aimed at influencing public opinion on divisive topics, and it typically leads to serious real-world
repercussions (
Barclay, 2018). Major digital platforms often come under criticism for not being clear about
how they make their decisions. They have a significant duty to manage content and reduce the spread of
harmful or misleading information (Gorwa, 2019). This lack of transparency can confuse users, as they
wonder why certain posts are taken down while other harmful content stays up, making it difficult to hold
anyone responsible (Roberts, 2019).
The lack of transparency in algorithmic content curation and moderation policies raises some serious
concerns about inconsistent enforcement and potential bias, especially during politically charged moments
(
Citron and Pasquale, 2019). While a few platforms have made strides to bring in third-party auditors and
share transparency reports, these efforts often fall short of fully explaining the processes behind content
moderation and appeals (
Gillespie, 2018). This gap highlights important ethical and legal dilemmas about
how to balance preventing harm with protecting free speech, as well as who really holds the responsibility
www.ijcrsee.com
829
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
for overseeing the vast expanse of online content (Suzor, 2019).
AI algorithms tend to favor content that ramps up user engagement, often leading to the amplifica-
tion of divisive or sensational topics, which only deepens the political divide. By shutting users off from a
variety of perspectives and reinforcing their existing biases, these AI-powered recommendation systems
create echo chambers and filter bubbles (Islam et al., 2024). Echo chambers are particularly common
on social media platforms, where users are mostly exposed to information that aligns with their beliefs.
This not only reinforces their deeply held convictions but also makes it increasingly difficult for them to
accept differing opinions (Jiang et al., 2021). Although the phrase “echo chamber” does not appear often
in political science literature, it is frequently used in the context of digital media studies. It illustrates how
individuals with like-minded opinions come together online, getting constant reinforcement that can skew
their understanding of reality and stifle genuine dialogue (Parry, 2006). Algorithmically personalized envi-
ronments, often referred to as “filter bubbles,” present significant challenges to democracy. These bubbles
create separate realities instead of fostering the shared truths necessary for informed participation in a
democratic society, as they limit users’ exposure to diverse opinions. Such personalization reinforces pre-
existing biases and desires by protecting people from information that challenges their beliefs, much like
subtle propaganda (Populismstudies, 2018).
Global data highlights troubling trends in how people perceive news bias and the spread of mislead-
ing information, contributing to what’s been dubbed the “infodemic.” Roughly 60% of people worldwide say
that most individuals only embrace information that backs up their own beliefs. This is particularly notice-
able in countries such as the US (68%), Turkey (69%), Serbia (70%), and Peru (71%). What’s intriguing is
that while 65% of respondents feel that people are actively hunting for opinions that reinforce their views,
only 34% admit to feeling stuck in an informational bubble (Konopliov, 2024). The power of this issue is
highlighted by the fact that over 70% of Europeans often encounter fake news (Watson, 2024).
The influence of fake news touches every corner of society, affecting how people, groups, and
governments respond to the false information that spreads like wildfire on social media. Much of this mis-
information is crafted to target specific demographics, aiming to stir up conflict and strengthen ideological
backing (Tandoc et al., 2018). In the US, political memes and viral videos are everywhere, often featuring
altered images, clips taken out of context, and even portraits created by AI. A couple of striking examples
include videos of Taylor Swift and Vice President Kamala Harris impersonating Donald Trump and seem-
ingly endorsing him (Bond, 2024).
AI has really changed the game when it comes to creating fake news. Large Language Models
(LLMs) can churn out a ton of readable and coherent content, thanks to their training on massive data-
sets. Plus, with rapid video generators like Sora, which can create detailed, Hollywood-style fake clips,
the spread and impact of misinformation have only grown (Virginia Tech, 2024). To tackle this problem, AI
models like Grover have been developed specifically to spot AI-generated fake news, and this approach
has shown to be quite effective (Zellers et al., 2019). It analyses the text and structure of articles to find
bias, inaccurate information, or other warning indicators using a combination of generation and detection
strategies. For separating machine-generated news from human-written news, Grover claims an accuracy
of above 92% (Gillham, 2025).
Discussion
The rapid advancement of artificial intelligence (AI) is changing politics and fundamentally altering
how democracies function in the modern world. With tools like generative models, deepfake, and innova-
tive content curation algorithms, we could see a surge in political participation, quicker communication,
and more tailored interactions between citizens and their leaders. As campaigns get better at fine-tuning
their messages, we could see a boost in voter turnout and overall democratic participation. Ideally, these
innovations could revitalize democratic processes by making political conversations more accessible,
dynamic, and aligned with the unique concerns of citizens.
The exciting potential of technology comes with some serious and complex risks that could un-
dermine social cohesion and the integrity of our democracy. The negative aspects of this technological
shift are starkly illustrated by deepfakes and AI-generated disinformation. Unlike simple text-based lies,
deepfakes create realistic videos of real individuals making false or misleading statements. These can be
www.ijcrsee.com
830
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
weaponized to attack opponents, suppress voter turnout, or increase social tensions, potentially desta-
bilizing entire electoral processes. The political implications are severe. This kind of manipulation poses
a significant threat to democracy, especially given how quickly and easily it can spread, as evidenced by
deepfake robocalls aimed at voters or viral videos that misrepresent politicians.
Beyond just deepfakes, the polarization and division in public discourse are getting worse thanks to
the widespread use of AI-driven content recommendation algorithms on social media. These algorithms
tend to create echo chambers and filter bubbles by prioritizing content that boosts user engagement,
which often means sensational, emotionally charged, or extreme ideological content. When users are cut
off from opposing viewpoints, it only reinforces confirmation bias and intensifies political tribalism. This
lack of common ground undermines the democratic dialogue and compromise we need, as it solidifies
partisan rifts and diminishes faith in democratic institutions. Moreover, the worldwide spread of false infor-
mation, often referred to as an “infodemic,” adds to the challenge by sowing doubt about which news and
information we can actually trust.
Furthermore, there’s a concerted effort to regulate AI and deepfakes. The European Union is tak-
ing the lead in this area, working on comprehensive regulatory frameworks that impose responsibility,
transparency, and content moderation on digital platforms. Key examples include the AI Act and the
Digital Services Act (DSA). These laws require the identification of AI-generated content, mandate risk
assessments, and ensure that harmful content is removed quickly, which are all crucial steps toward pro-
tecting democratic discourse. Despite the complex challenges at play, it’s hard to say just how effective
these restrictions will be. To begin with, politicians often struggle to keep pace with the rapid evolution of
technology. On top of that, the global nature of the internet complicates jurisdiction enforcement, and the
ongoing debates about free speech, censorship, and platform responsibility create tough moral dilemmas.
Critics warn that overly broad regulations might hinder free expression, while too little oversight could al-
low misinformation to thrive.
On a more profound level, however, the phenomenon The Washington Post refers to as the “liar’s
dividend” is the increasingly prevalent practice of politicians claiming that compromising statements or vid-
eos were produced artificially (either by deepfake technology or other AI tools) when there is no supporting
evidence (Verma and De Vynck, 2024). This approach sets a concerning precedent by leveraging uncer-
tainty around the accuracy of digital content to dodge accountability. It muddles the line between manipula-
tion and the actual truth. This trend sparks major concerns regarding democracy and social responsibility,
as it diminishes the trust that is so important for meaningful discussions in the public sphere. Even when
the information is accurate, people struggle to make informed choices if everything is labelled as potentially
fake. This also gives those in power the ability to alter the truth without facing any consequences.
Keeping up with technological advancements is crucial in the battle against AI-driven misinforma-
tion. AI detection tools are getting better at spotting and flagging fake media. We really need to boost
public awareness about AI-generated content and foster critical thinking to help people resist manipula-
tion. Additionally, by putting transparency measures in place, like better content moderation disclosures,
we can build more public trust and accountability.
The way societies tackle the ethical, legal, and technological challenges posed by AI will ultimately
shape its impact on political communication and the integrity of democracy. We need to encourage col-
laborative governance frameworks that bring together governments, tech companies, academia, civil so-
ciety, and the public to set guidelines and standards for ethical AI use. To find the right balance between
fostering innovation and upholding essential democratic values like accountability, transparency, freedom
of expression, and human rights, we need a strategy that involves multiple stakeholders. Rather than
diminishing democratic participation, AI has the potential to enhance it, provided we integrate ethical
considerations and maintain strict oversight in its development and application.
Conclusion
In conclusion, we must acknowledge AI’s impressive potential to transform political communication
and improve democratic engagement. However, we also need to be mindful of the serious threats posed
by deepfakes and AI-driven misinformation campaigns to the future of this technology. It’s essential that
we take prompt action to combat the erosion of trust in information, the increasing social divides, and the
www.ijcrsee.com
831
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
flaws in our electoral systems. Strong laws, advanced technology, public awareness, and global coopera-
tion are vital for democracies to mitigate these risks and safeguard the integrity of their political institu-
tions. The decisions we make today about AI regulation will either uphold the foundational principles of
democracy or allow for manipulative technology that diminishes informed public engagement.
It’s crucial for future research to dive deeper into interdisciplinary solutions that blend ethical AI
design, media literacy education, and inclusive governance frameworks. Scholars, educators, policymak-
ers, and tech developers need to join forces to create evidence-based strategies that strengthen civic
resilience and safeguard democratic discourse. By nurturing a culture of transparency, critical thinking,
and accountability in technology, we can ensure that AI serves as a tool for empowering democracy in-
stead of causing disruption. The relationship between AI and politics is quite complex, and it requires our
continuous joint efforts. We need to work together to nurture the intellectual and democratic development
of future generations.
Acknowledgement
The paper presents findings of a study developed as a part of the research project “Serbia and
challenges in international relations in 2025”, financed by the Ministry of Science, Technological Develop-
ment and Innovation of the Republic of Serbia, and conducted by Institute of International Politics and
Economics, Belgrade during year 2025.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial,
or not-for-profit sectors.
Conflict of interests
The authors declare no conflict of interest.
Data availability statement
The original contributions presented in the study are included in the article/supplementary material,
further inquiries can be directed to the corresponding author/s.
Institutional Review Board Statement
Not applicable.
Author Contributions
Conceptualization, B.R., V.L. and B.I.; methodology, B.R.; formal analysis, B.R. and B.I.; writing—
original draft preparation, B.R. and V.L.; writing—review and editing, B.R. and B.I. All authors have read
and agreed to the published version of the manuscript.
References
Alvarez, R. M., Eberhardt, F., & Linegar, M. (2023, July). Generative AI and the Future of Elections
(CSSPP White Paper).
California Institute of Technology Center for Science, Society, and Public Policy. https://lindeinstitute.caltech.edu/docu-
ments/25475/CSSPP_white_paper.pdf
Appel, M., & Prietzel, F. (2022). The detection of political deepfakes. Journal of Computer-Mediated Communication, 27(4),
zmac008. http://dx.doi.org/10.1093/jcmc/zmac008
Baltezarević, V., Baltezarević, R., & Milovanović, S. (2014). Between the lines and through the images. Informatologija, 47(1),
29-35. https://hrcak.srce.hr/le/178309
Baltezarević, R., Baltezarević, B., Kwiatek, P., & Baltezarević, V. (2019). The impact of virtual communities on cultural identity.
Symposion, 6(1), 7-22. https://doi.org/10.5840/symposion2019611
Baltezarević, R., & Baltezarević, I. (2024). Students’ Attitudes on The Role of Articial Intelligence (Ai) In Personalized Learn-
ing. International Journal of Cognitive Research in Science, Engineering and Education (IJCRSEE), 12(2), 387-397.
http://dx.doi.org/10.23947/2334-8496-2024-12-2-387-397
www.ijcrsee.com
832
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
Bareis, J., & Katzenbach, C. (2022). Talking AI into being: The narratives and imaginaries of national AI strategies and their
per-
formative politics. Science, Technology, & Human Values, 47(5), 855-881. https://doi.org/10.1177/01622439211030007
Barclay D. A. (2018). Fake news, propaganda, and plain old lies: how to nd trustworthy information in the digital age. Lanham,
MD: Rowman & Littleeld.
Battista, D., & Uva, G. (2023). Exploring the Legal Regulation of Social Media in Europe: A Review of Dynamics and
Chal-
lenges—Current Trends and Future Developments. Sustainability, 15(5), 4144. https://doi.org/10.3390/su15054144
Bessi, A., & Ferrara, E. (2016). Social Bots Distort the 2016 US Presidential Election Online Discussion. SSRN Electronic
Journal. https://ssrn.com/abstract=2982233
Bond, S. (2024, December). How AI deepfakes polluted elections in 2024. NPR. https://www.npr.org/2024/12/21/nx-s1-
5220301/deepfakes-memes-articial-intelligence-elections
Brady, W. J., Wills, J. A., Jost, J. T., Tucker, J. A., & Van Bavel, J. J. (2017). Emotion shapes the diffusion of moralized content
in social networks. Proceedings of the National Academy of Sciences, 114(28), 7313–7318. http://dx.doi.org/10.1073/
pnas.1618923114
Bradshaw, S., & Howard, P. N. (2018). Challenging Truth and Trust: A Global Inventory of Organized Social Media Manipula-
tion. Oxford University.
Brannon, W., Beeferman, D., Jiang, H., Heyward, A., & Roy, D. (2024). AudienceView: AI-assisted interpretation of audience
feedback in journalism. arXiv. https://arxiv.org/abs/2407.12613
Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J.D., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A. &
Agarwal, S. (2020). Language models are few-shot earners. Advances in Neural Information Processing Systems, 33,
1877-1901. https://arxiv.org/abs/2005.14165
BSI (2025). Digression: Social Bots and Chatbots.
Bundesamt für Sicherheit in der Informationstechnik. https://www.bsi.bund.
de/EN/Themen/Verbraucherinnen-und-Verbraucher/Informationen-und-Empfehlungen/Onlinekommunikation/Soziale-
Netzwerke/Sichere-Verwendung/Exkurs-bots/social-bots.html
Burkhardt, J. M. (2017). History of Fake News. Library Technology Reports, 53(8), 5-9. https://journals.ala.org/index.php/ltr/
article/view/6497/8631
Carr, N. G. (2011). The shallows: What the internet is doing to our brains. W. W. Norton & Company.
CBS News. (2019, May). Doctored Nancy Pelosi video highlights threat of “deepfake” tech.
https://www.cbsnews.com/news/
doctored-nancy-pelosi-video-highlights-threat-of-deepfake-tech-2019-05-25/
Chang, T. (2025, April). The AI Bot Epidemic: The Imperva 2025 Bad Bot Report. Thales Group.https://cpl.thalesgroup.com/
blog/access-management/ai-bots-internet-trafc-imperva-2025-report
Chawla, R. (2019). Deepfakes: How a pervert shook the world. International Journal of Advance Research and Development,
4(6), 4–8. https://www.semanticscholar.org/paper/Deepfakes-%3A-How-a-pervert-shook-the-world-Chawla/c3b3a6d-
27dbbfed4df630b39fc0a8a6692b1828a
Cheguri, P. (2023). Deepfake Technology: Concerns Raised in the Advertising Industries.
Analytics Insight. https://www.analyt-
icsinsight.net/topic/deepfake-technology
Chester, J., & Montgomery, K. C. (2017). The role of digital marketing in political campaigns. Internet Policy Review, 6(4), 1–20.
http://dx.doi.org/10.14763/2017.4.773
Citron, D. K., & Pasquale, F. (2019). The scored society: Due process for automated predictions. Washington Law Review,
89(1), 1–33. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2376209
Coeckelbergh, M. (2022). The Political Philosophy of AI: An Introduction. John Wiley & Sons: New York, NY, USA.
Corney, D., Wilkinson, K., & Cann, R. (2024, June). The AI election: How Full Fact is leveraging new technology for UK
gen-
eral election fact checking. Full Fact. https://fullfact.org/blog/2024/jun/the-ai-election-how-full-fact-is-leveraging-new-
technology-for-uk-general-election-fact-checking/
Crawford, K. (2021). The atlas of AI: Power, politics, and the planetary costs of articial intelligence. Yale University Press.
https://doi.org/10.12987/9780300252392
Crilley, R. (2018). International relations in the age of ‘post-truth’ politics. International Affairs, 94(2), 417-425.
http://dx.doi.
org/10.1093/ia/iiy038
Cruz, B. (2024, August). 2024 Deepfakes Guide and Statistics. Security. https://www.security.org/resources/deepfake-statistics/
Dudeld, A. (2025, August). How to stop AI chatbots going rogue. Full Facts. https://fullfact.org/technology/how-to-stop-ai-
chatbots-going-rogue/
Ethayarajh, K., Xu, W., Muennighoff, N., Jurafsky, D., & Kiela, D. (2024). KTO: Model Alignment as Prospect Theoretic Optimi-
zation. arXiv. https://arxiv.org/abs/2402.01306
European Commission. (2021, April). Proposal for a Regulation of the European Parliament and of the Council Laying Down
Harmonised Rules on Articial Intelligence (Articial Intelligence Act) and amending certain union legislative acts.
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52021PC0206
European Parliament. (2022, July). Digital Services: landmark rules adopted for a safer, open online environment. https://www.eu-
www.ijcrsee.com
833
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
roparl.europa.eu/news/en/press-room/20220701IPR34364/digital-services-act-eu-rules-to-make-digital-platforms-safer
Factiverse. (2025, May). How Factiverse Scans the Web to Tackle Misinformation at Scale. https://www.factiverse.ai/blog/how-
factiverse-scans-the-web-to-tackle-misinformation-at-scale
Ferrara, E., Varol, O., Davis, C., Menczer, F., & Flammini, A. (2016). The rise of social bots. Communications of the ACM, 59(7),
96–104. https://dl.acm.org/doi/10.1145/2818717
Fetzer, J.H., & Fetzer, J.H. (1990). What Is Articial Intelligence? Springer.
Fullfact. (2025). Find and Fight Bad Information.
https://fullfact.ai/about/
Funk, A., Shahbaz, A., & Vesteinsson, K. (2023, November). The Repressive Power of Articial Intelligence. Freedom House.
https://freedomhouse.org/report/freedom-net/2023/repressive-power-articial-intelligence
Gallo, M., Fenza, G., & Battista, D. (2022). Information Disorder: What about global security implications? Rivista di Digital
Politics, 2(3), 523-538. https://doi.org/10.53227/106458
Gilbert, D. (2025, June). AI Chatbots Are Making LA Protest Disinformation Worse. Wired.
https://www.wired.com/story/grok-
chatgpt-ai-los-angeles-protest-disinformation
Gillham, J. (2025, September). Grover AI Content Detection Review. Originality.ai https://originality.ai/blog/grover-ai-content-
detection-review
Gillespie, T. (2018). Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social
Media. Yale University Press.
Gorwa, R. (2019). The platform governance triangle: conceptualising the informal regulation of online content. Internet Policy
Review, 8(2). https://doi.org/10.14763/2019.2.1407
Guess, A.M., Lerner, M., Lyons, B., Montgomery, J.M., Nyhan, B., Reier, J., & Sircar, N. (2020). A digital media literacy intervention
increases discernment between mainstream and false news in the United States and India. Proceedings of the National
Academy of Sciences of the United States of America, 117(27), 15536-15545. https://doi.org/10.1073/pnas.1920498117
Hagerty, A., & Rubinov, I. (2019, July). Global AI ethics: A review of the social impacts and ethical implications of articial intel-
ligence. arXiv. https://arxiv.org/abs/1907.07892
Hetrick. C. (2024, July). How to spot AI fake news – and what policymakers can do to help. USC Price School of Public Policy.
https://priceschool.usc.edu/news/ai-election-disinformation-biden-california-europe/
Hu, L., Wei, S., Zhao, Z., & Wu, B. (2022). Deep learning for fake news detection: A compre-hensive survey. AI Open, 3, 133-
155, https://doi.org/10.1016/j.aiopen.2022.09.001
Islam, M.B.E., Haseeb, M., Batool, H., Ahtasham, N., & Muhammad, Z. (2024). AI Threats to Politics, Elections, and Democ-
racy: A Blockchain-Based Deepfake Authenticity Verication Framework. Blockchains 2024, 2(4), 458–481. https://doi.
org/10.3390/blockchains2040020
Jiang, J., Ren, X., & Ferrara, E. (2021). Social Media Polarization and Echo Chambers in the Context of COVID-19: Case
Study. JMIRx Med, 2(3), e29570 https://doi.org/10.2196/29570
Kerly, R. (2020, August). How Deepfakes Are Changing Digital Marketing. Loop Digital. https://www.loop-digital.co.uk/the-rise-
of-deepfake-technology/
Khare, Y. (2023, April). The Role of AI in Political Campaigns: Revolutionizing the Game. Analytics Vidhya. https://www.analyt-
icsvidhya.com/blog/2023/04/the-role-of-ai-in-political-campaigns-revolutionizing-the-game/
Klinger, U., Kreiss, D., & Mutsvairo, B. (2023). Platforms, Power, and Politics: A Model for an Ever-changing Field. Political
Communication Report, 27, 1-6. http://dx.doi.org/10.17169/refubium-39045
Konopliov, A. (2024, June). Key Statistics on Fake News & Misinformation in Media in 2024. Redline Digital. https://redline.
digital/fake-news-statistics/
Korshunov, P., & Marcel, S. (2018, December). DeepFakes: a New Threat to Face Recognition? Assessment and Detection.
arXiv. https://arxiv.org/abs/1812.08685
Kumar, A., & Garg, G. (2020). Systematic Literature Review on Context-Based Sentiment Analysis in Social Multimedia. Multi-
media Tools and Applications, 79(21-22), 15349–15380. https://doi.org/10.1007/s11042-019-7346-5
Kumar, S. (2025, May). Why Governments Worldwide Are Enacting Stricter AI Deepfake Regulations in 2025. Medium.
https://medium.com/@meisshaily/why-governments-worldwide-are-enacting-stricter-ai-deepfake-regulations-in-
2025-32a61309366c
Landrin, S. (2024, May). India’s general election is being impacted by deepfakes. LeMonde. https://www.lemonde.fr/en/pixels/
article/2024/05/21/india-s-general-election-is-being-impacted-by-deepfakes_6672168_13.html
Lazaro Cabrera, L. (2024, May). AI Policy & Governance, European Policy, Free Expression. EU AI Act Brief – Pt. 3, Freedom
of Expression. Center for Democracy & Technology. https://cdt.org/insights/eu-ai-act-brief-pt-3-freedom-of-expression/
Loth, A., Kappes, M., & Pahl, M.-O. (2024, April). Blessing or curse? A survey on the Impact of Generative AI on Fake News.
arXiv. https://arxiv.org/abs/2404.03021
Lutkevich, B. & Hildreth, S. (2022, February). Social listening (social media listening). TechTarget. https://www.techtarget.com/
searchcustomerexperience/denition/social-media-listening
www.ijcrsee.com
834
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
Maheshwari, S. (2024, August). Brands Love Inuencers (Until Politics Get Involved). The New York Times.
https://www.ny-
times.com/2024/08/12/business/media/inuencers-politics-ai-analysis.html
Matheis, A. (2023). How can articial intelligence be used in political communication? Wegewerk. https://www.wegewerk.com/
en/blog/how-can-articial-intelligence-be-used-in-political-communication/
Mermoud, A. (2017, August). The Power of Big Data and Psychographics in politics. Swiss Intell. https://swissintell.ch/the-
power-of-big-data-and-psychographics-in-politics/
Murali, P., Hernandez, J., McDuff, D., Rowan, K., Suh, J., & Czerwinski, M. (2021, January). AffectiveSpotlight: Facilitating the commu-
nication of affective responses from audience members during online presentations. arXiv. https://arxiv.org/abs/2101.12284
NBC News (2023, December). Putin quizzed about AI and body doubles by his apparent deepfake. https://www.nbcnews.com/
video/putin-quizzed-about-ai-and-body-doubles-by-his-apparent-deepfake-200210501620
Nida-Rumelin, J., & Weidenfeld, N. (2019). Umanesimo digitale: un’etica per l’epoca dell’Intelligenza articiale. FrancoAngeli.
Noto, G. (2024, May). Scammers siphon $25M from engineering rm Arup via AI deepfake ‘CFO’. CFO Dive.
https://www.
cfodive.com/news/scammers-siphon-25m-engineering-rm-arup-deepfake-cfo-ai/716501/
Nunziata, F. (2021). Il platform leader. Rivista di Digital Politics, 1(1), 127-146.
https://www.rivisteweb.it/doi/10.53227/101176
Obot, O. U., Attai, K. F., Onwodi, G. O., James, I., & John, A. (2025). Sentiment analysis of electronic word of mouth (E-
WoM) on e-learning. In M. Khosrow-Pour (Ed.), Encyclopedia of Information Science and Technology (6th ed., ch. 57).
IGI Global. https://doi.org/10.4018/978-1-6684-7366-5.ch057
Pamment, J., Nothhaft, H., & Fjällhed, A. (2018). Countering information inuence activities: A handbook for communicators.
Swedish Civil Contingencies Agency (MSB). https://rib.msb.se/ler/pdf/28697.pdf
Parry, R. (2006, December). The GOP’s $3 Bn Propaganda Organ. The Baltimore Chronicle.
https://baltimorechronicle.com/
Pennycook, G., Cannon, T.D., & Rand, D.G. (2018). Prior exposure increases perceived accuracy of fake news. Journal of
Experimental Psychology: General, 147(12), 1865-1880.
https://doi.org/10.1037/xge0000465
Petrosyan, A. (2025, May). Potential inuence of AI and deepfakes on upcoming elections 2024, by country. Statista. https://
www.statista.com/statistics/1534957/global-potential-inuence-ai-elections-by-country/
Political Communication. (2023). AI and Political Communication. Political Communication Report
, 27(Spring). https://political-
communication.org/article/ai-and-political-communication/
Populismstudies. (2018). Filter Bubbles. ECPS.
https://www.populismstudies.org/Vocabulary/lter-bubbles/
Rajashekhargouda, P. (2022). Sentimental Analysis on Amazon Reviews Using Machine Learning. In Karuppusamy, P.,
García
Márquez, F.P., & Nguyen, T.N., (Eds.). Ubiquitous Intelligent Systems (pp. 467–477). Springer Nature Singapore.
Roberts, S. T. (2019). Behind the screen: Content moderation in the shadows of social media. Yale University Press.
Roozenbeek, J., & van der Linden, S. (2020). Breaking Harmony Square: A game that “inoculates” against political
misinforma-
tion. Harvard Kennedy School (HKS) Misinformation Review, 1(8). https://doi.org/10.37016/mr-2020-47
Samoili, S., López Cobo, M., Gómez, E., De Prato, G., Martínez-Plumed, F., & Delipetrev, B., (2020). AI watch. Dening
arti-
cial intelligence. Towards an operational denition and taxonomy of articial intelligence (EUR 30117 EN). Publications
Ofce of the European Union. https://doi.org/10.2760/382730
Schneider, B. (2017, June). How Vote Leave Used Data Science and A/B Testing to Achieve Brexit. AB Tasty. https://www.
abtasty.com/blog/data-science-ab-testing-vote-brexit/
Seitz-Wald, A. (2024, February). A New Orleans magician says a Democratic operative paid him to make the fake Biden
robocall. NBC News. https://www.nbcnews.com/politics/2024-election/biden-robocall-new-hampshire-strategist-rc-
na139760?_ga=2.181210351.976717714.1719011557-176973521.1719011550
Shen, T., Ruixian, L., Ju, B., & Zheng, L. (2018). ‘Deep Fakes’ Using Generative Adversarial Networks (GAN)
(Report No. 16).
Noiselab, University of California, San Diego. http://noiselab.ucsd.edu/ECE228_2018/Reports/Report16.pdf
Somers, M. (2020, July). Deepfakes, explained.
MIT Sloan School of Management. https://mitsloan.mit.edu/ideas-made-to-
matter/deepfakes-explained
Statista Research Department. (2024). U.S. adults worry about AI-generated political propaganda 2023. Statista. https://www.
statista.com/statistics/1471069/us-adults-ai-generated-political-propaganda/
Surfshark. (2025). Deepfake statistics in early 2025: how frequently are famous people targeted? https://surfshark.com/re-
search/study/deepfake-statistics
Sustainability Directory. (2025, May). How Effective Is Digital Literacy in Addressing Misinformation? https://sustainability-
directory.com/question/how-effective-is-digital-literacy-in-addressing-misinformation/
Suzor, N. (2019). Lawless: The secret rules that govern our digital lives. Cambridge University Press. https://doi.
org/10.1017/9781108666428
Tandoc, E. C., J., Lim, Z. W., & Ling, R. (2018). Dening “Fake News”: A typology of scholarly denitions. Digital Journalism,
6(2), 137–153. https://doi.org/10.1080/21670811.2017.1360143
Thornhill, J. (2024, June). The danger of deepfakes is not what you think. The Straits Times. https://www.straitstimes.com/
opinion/the-danger-of-deepfakes-is-not-what-you-think
www.ijcrsee.com
835
Baltezarević, R., Lović, V., & Baltezarević, I. (2025). Between Progress and Peril: The Role of Artificial Intelligence (AI) in
Shaping Modern Political Communication, International Journal of Cognitive Research in Science, Engineering and Education
(IJCRSEE), 13(3), 823-835.
Verma, P., & De Vynck, G. (2024, January). AI is destabilizing ‘the concept of truth itself’ in 2024 election. The Washington
Post. https://www.washingtonpost.com/technology/2024/01/22/ai-deepfake-elections-politicians
Virginia Tech. (2024, February). AI and the spread of fake news sites: Experts explain how to counteract them. Virginia Tech
News. https://news.vt.edu/articles/2024/02/AI-generated-fake-news-experts.html
Viudes, F. J. (2023). Revolucionando la política: El papel omnipresente de la IA en la segmentación y el targeting de campañas
modernas. Más poder local, (53), 146-151. https://doi.org/10.56151/maspoderlocal.183
Vosoughi, S., Deb, R. & Aral, S. (2018). The spread of true and false news online. Science, 359(6380),1146-1151. https://www.
science.org/doi/10.1126/science.aap9559
Wakeeld, J. (2019, November). Brittany Kaiser calls for Facebook political ad ban at Web Summit. BBC News. https://www.
bbc.com/news/technology-50234144
Watson, A. (2024, January). Fake news in Europe - statistics & facts. Statista. https://www.statista.com/topics/5833/fake-
news-in-europe/#topicOverview
Weforum (2024). Fake news undermines democracy, warns global survey. https://www.weforum.org/videos/inuence-of-fake-news/
Westerlund, M. (2019). The emergence of deepfake technology: A review. Technology Innovation Management Review, 9(11),
40-53. https://doi.org/10.22215/timreview/1282
Yu, C. (2024). How will AI steal our elections? (OSF Preprint, un7ev). Center for Open Science. https://doi.org/10.31219/osf.io/un7ev
Zandt, F. (2024, March). How Dangerous are Deepfakes and Other AI-Powered Fraud? Statista Daily Data. https://www.
statista.com/chart/31901/countries-per-region-with-biggest-increases-in-deepfake-specic-fraud-cases/
Zannettou, S., Sirivianos, M., Blackburn, J. & Kourtellis, N. (2019). The Web of False Information: Rumors, Fake News,
Hoaxes, Clickbait, and Various Other Shenanigans. Journal of Data and Information Quality, 1(3), Article No. 10. https://
doi.org/10.1145/3309699
Zellers, R., Holtzman, A., Rashkin, H., Bisk, Y., Farhadi, A., Roesner, F., & Choi, Y. (2019). Defending against neural fake news.
Advances in neural information processing systems, 32. https://arxiv.org/abs/1905.12616
